Political Ideology and Racism

Over at the Volokh conspiracy, contributor James Lindgren (who is quite adept with statistical analysis software and the data at issue) contends that white Democrats are significantly more racist than are white Republicans:

Eugene,

To your list I would add that, over the last 30 years, typical white Republicans in the general public have been significantly less traditionally racist than typical white Democrats.

For example, in the 1996-2004 General Social Surveys, 11.9% of white Democrats think that differences in black/white success are due to black inborn disability, compared to only 9.1% of white Republicans. Similarly, 14.5% of white Democrats favor laws against racial intermarriage, compared to 12.2% of white Republicans.

In the two decades before 1996, these differences tend to be much stronger. Republicans have traditionally been less traditionally racist.

So, although there is a shift in traditional racism toward members of the Republican Party, members of the Democratic Party have been still more racist over the last decade--though the Republican Party may overtake it soon. Only if one uses "modern racism" measures that conflate opposition to big government with racism do Republicans usually score as significantly more racist than Democrats.

Given the tiny percentages at issue, it is astonishing that Lindgren would try to assert significant racism or lack thereof on the basis of his recited figures. But beyond that, if the data truly supports his claims, why is he the only one making these claims? Would Karl Rove truly sit on such explosive data, as the Republicans attempt to make headway among this nation's minority populations?

Although I don't have access to the same tools as Lindgren, it is possible to perform a basic reality check on his assertion through Berkeley's SDA Archive site. Using the "Frequencies or crosstabulation" tool, looking at the General Social Surveys, 1972-2004 Cumulative File, focusing on Lindgren's hallmark of racism, the question of inborn ability [RACDIF4], limiting results to white respondants [RACE(1)], and weighing results with a non-response adjustment [WT2004NR], I see the following results:



When I look at expression of admiration for blacks [ADMIRBLK], I see the following results:



By Lindgren's measure (and let me again emphasize that I disgree with his measure), if Democrats can be said to be significantly more racist than Republicans, conservatives (and moderates) would appear to be vastly more racist than liberals.

When I look at Lindgren's argument that Democrats are more likely to favor laws against interracial marriage [RACMAR], he's right - the survey indicates that strong and moderate Democrats and strong Republicans are more likely to support such laws:



But when liberal and conservative ideology is examined, it seems that strong support for such laws comes from people who describe themselves as moderate, polically conservative and strongly conservative, while political liberals are significantly less likely to support such laws:



Perhaps the difference between the two results is explained by the fact that political moderates are more likely to self-identify as Democrats than they are as Republicans:



I don't agree with Lindgren's extrapolations from the data - I don't think his data supports his claims, and I think he knows it. I know that at least one statistics guru passes by this blog on occasion - I would love to hear an expert opinion on this, as a final word on my clumsy analysis as presented above.